dsh-voice: Adding Full-Duplex Voice Mode to DeepSeek Harness
`haoku123/dsh-voice` is a full-duplex voice plugin for DeepSeek Harness (DSH), licensed under MIT, designed to provide real-time voice interaction experiences for web clients. Its core pipeline follows the flow of "streaming ASR → LLM → TTS with interruption support." The browser side captures Raw PCM audio, while the host side leverages `SenseVoice` (based on `sherpa-onnx`) for Chinese speech recognition, punctuation prediction, and text normalization, and uses `msedge-tts` to stream audio and captions back to the client. The plugin supports three input methods: tap-to-talk for continuous dictation, hold-to-record via the send key or microphone (swipe up to discard), and keyboard-free dictation by holding the Ctrl key (Esc to discard). It also features real-time caption preview.
Read MoreOpenAI Integrates GPT-Live Full-Duplex Voice Access into Codex, Paving the Way for Hands-Free Operation of Coding Agents
On July 23, 2026, OpenAI integrated GPT-Live full-duplex voice access into Codex and ChatGPT Work in ChatGPT desktop build 26.715. Developers can use natural voice to initiate parallel coding tasks across threads between Chat, Work and Codex, review PRs, and remotely guide Agents. macOS also supports Appshots screen context. Based on OpenAI's official Changelog and Voice documentation, this article sorts out the GPT-Live architecture, typical scenarios, enabling steps, quota and permission boundaries, to help readers understand the interactive upgrade of Agentic Coding from typing commands to hands-free orchestration.
Read More